The deterioration of trabecular microarchitecture induced by elevated bone turnover is increasingly recognized as a factor in the pathogenesis of osteoporotic fractures. We investigated the effect of the reduction of turnover with risedronate on trabecular architecture in postmenopausal women with osteoporosis. Iliac crest bone biopsy specimens taken before and after 3 years of treatment from patients receiving risedronate 5 mg daily (n = 21) or placebo (n = 17) were analyzed using 3-D microcomputed tomography. We found a significant correlation between baseline bone turnover and bone loss in the placebo group, providing evidence that higher turnover induced higher bone loss leading to a greater degree of architectural degradation. When patients were classified into two groups based on baseline bone turnover (MS/BS less than or greater than the median value for the entire cohort), significant decreases in trabecular bone volume (BV/TV, P = 0.009) and trabecular thickness (Tb.Th*, P = 0.008) and an increase in marrow star volume (Ma.St.V, P = 0.008), a measure of trabecular porosity, were observed in the higher turnover (MS/BS> median) placebo-treated patients. The trabecular structure shifted from plates to rods as shown by an increase in structure model index (SMI, P = 0.028) and bone surface to bone volume ratio (BS/BV, P = 0.006). The changes from baseline in the lower turnover (MS/BS<median) placebo patients were variable and not statistically significant. In the risedronate group, the bone volume and the architectural parameters did not change significantly from baseline values in either the higher or the lower turnover groups. Comparing the pair-wise changes from baseline in the higher turnover group, the placebo group experienced decreases in BV/TV (P = 0.071) and Tb.Th* (P = 0.012), and increase in Ma.St.V (P = 0.043), compared to the risedronate-treated women. Also, in comparison to the risedronate group, the trabecular structures in the placebo group were more rod-like, indicated by higher SMI (P = 0.009) and BS/BV (P = 0.02). The results demonstrated that trabecular architecture deteriorated significantly in the placebo-treated women who had higher bone turnover at baseline, and this deterioration was prevented by 3 years of risedronate treatment, presumably because of the reduction in bone turnover. The preservation of architecture may be a contributory mechanism by which risedronate reduces the risk of vertebral fractures in osteoporotic women.